Chunk extensions

[Lukasa]

Well, you predicted it from #19! And here it is.

Right now, amazingly, urllib3 actually tests that it can handle chunk extensions. This is the first I have seen that demonstrates that we actually do it, but here we are.

I am open to being told that we should drop that support, but I wanted to ask whether you think it's worth doing either of the following two things:

1. Fully supporting chunk extensions.
2. Merely tolerating them (ignoring them if they are received)

[njsmith]

Currently h11 tolerates chunk extensions (parses their edges, then discards the contents). Does urllib3 actually have an API to expose these to the user currently?

…

On Dec 7, 2016 9:29 AM, "Cory Benfield" ***@***.***> wrote: Well, you predicted it from #19 <https://github.com/njsmith/h11/issues/19>! And here it is. Right now, amazingly, urllib3 actually tests that it can handle chunk extensions. This is the first I have seen that demonstrates that we actually do it, but here we are. I am *open* to being told that we should drop that support, but I wanted to ask whether you think it's worth doing either of the following two things: 1. Fully supporting chunk extensions. 2. Merely *tolerating* them (ignoring them if they are received) — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <https://github.com/njsmith/h11/issues/26>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAlOaE8iBZfbfcpg0NzKDTur20FXisnkks5rFuztgaJpZM4LG3td> .

[Lukasa]

No, but I have a test that is failing that sends them. I assumed this was merely a lack of tolerance on h11's part, but I'll double-check and get back to you: presumably that test is failing for some other reason.

[Lukasa]

Ah, ok, nevermind: this is an artefact of the way the tests work.

[njsmith]

Best kind of bug report :-)